A Method for Analyzing Compressor Aerodynamic Uncertainties Based on Nested High-Dimensional Models

Abstract

This invention relates to the field of compressor technology and discloses a method for analyzing compressor aerodynamic uncertainties based on a nested high-dimensional model representation. The method includes the following steps: constructing a compressor aerodynamic response function and establishing a mapping relationship between the uncertainty parameters and compressor aerodynamic performance indicators; decomposing the mapping relationship based on a high-dimensional model representation method to obtain multiple component functions; recombining the component functions in a nested manner according to parameter categories to construct a nested high-dimensional model representation model; performing compressor aerodynamic uncertainty analysis based on the nested high-dimensional model representation model; and outputting the analysis results, which are used to guide compressor manufacturing tolerance allocation, geometric optimization design, or operational condition evaluation. This invention provides a compressor aerodynamic uncertainty analysis method based on a nested high-dimensional model representation, which can improve the engineering interpretability of the results.

Description

Technical Field

[0001] This invention relates to the field of compressor technology, and in particular to a method for analyzing compressor aerodynamic uncertainties based on nested high-dimensional model representation. Background Technology

[0002] As a key component in aero-engines and energy systems, the aerodynamic performance of the compressor directly affects the overall efficiency, stability, and safety of the machine. During the design, manufacturing, and operation of compressors, geometric uncertainties inevitably arise due to machining errors, assembly deviations, and changes in operating conditions. These uncertainties can cause aerodynamic performance to deviate from design targets, severely restricting the engineering application of high-efficiency compressors.

[0003] To analyze the impact of uncertainties on compressor performance, uncertainty analysis methods are commonly used in engineering. However, due to the large number of compressor geometric parameters, the complexity of parameter types, and the significant coupling relationships between parameters, traditional analysis methods based on single parameters or low-dimensional parameters are insufficient to reveal the comprehensive impact of different types of parameters and their interactions on aerodynamic performance.

[0004] High-dimensional model representation (HDMR) methods reduce computational complexity to some extent by decomposing high-dimensional problems into several low-dimensional components and can be used for uncertainty analysis. However, existing HDMR methods typically analyze the impact of uncertainty only at the "parameter dimension" level, lacking a systematic expression of the contributions of different parameter categories and their coupling mechanisms, resulting in shortcomings in the engineering physics interpretation and design decision support of the analysis results. Existing compressor uncertainty analysis methods mainly include statistical analysis methods based on Monte Carlo simulations and sensitivity analysis methods based on surrogate models. Some studies have introduced HDMR methods to decompose high-dimensional parameters, but still only use HDMR as a surrogate model or as the analysis object for multi-parameter pairs, without fully utilizing its hierarchical and inherent interpretable structure.

[0005] The above methods have the following shortcomings: it is difficult to quantitatively analyze the sources of uncertainty at the parameter category level; it is difficult to reveal the interaction effects between different categories of parameters; the analysis results are mostly at the mathematical and statistical level, with insufficient engineering interpretability; and it is not conducive to providing clear decision-making basis for manufacturing tolerance control and structural optimization. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a method for analyzing compressor aerodynamic uncertainties based on nested high-dimensional model representation, which can improve the engineering interpretability of the results.

[0007] This invention provides a method for analyzing compressor aerodynamic uncertainties based on nested high-dimensional model representation, including the following steps: acquiring and classifying the uncertainty parameters of the compressor, and dividing the uncertainty parameters into multiple parameter categories with different engineering physical meanings; A compressor aerodynamic response function is constructed to establish a mapping relationship between uncertain parameters and compressor aerodynamic performance indicators. The mapping relationship is then decomposed based on a high-dimensional model representation method to obtain multiple component functions. Based on the parameter categories, the component functions obtained from the decomposition are reorganized in a nested manner to construct a nested high-dimensional model representation model; The uncertainty parameters are input into the nested high-dimensional model representation model to perform compressor aerodynamic uncertainty analysis; The output analysis results are used to guide the allocation of manufacturing tolerances, geometric optimization design, or operational condition evaluation of the compressor.

[0008] Specifically, the mapping relationship between uncertain parameters and compressor aerodynamic performance indicators is established, and the mapping relationship is decomposed based on a high-dimensional model representation method, including: , in, For zero-order component terms, It is a first-order component term. Representing variables and The second-order component term generated by the coupling of related effects after removing the influence of individual effects.

[0009] Specifically, uncertain parameters include geometric dimension parameters, profile control parameters, angle parameters, material property parameters, incoming flow condition parameters, and operating condition parameters; the parameter categories can be divided by: uncertainty source, geometric dimension, structural region, aerodynamic component, or blade type.

[0010] Specifically, aerodynamic performance indicators include: adiabatic efficiency, isentropic efficiency, total pressure ratio, static pressure ratio, flow rate, power, power consumption, surge margin, blockage margin, loss coefficient, pressure recovery coefficient, and other indicators characterizing the aerodynamic performance of the compressor; the aerodynamic response function is constructed by combining numerical simulation or experimental testing with a high-dimensional model.

[0011] Specifically, nested restructuring refers to: Let there be a total of The parameter category, the first The categories include Given several uncertain parameters, the second-order expansion of the nested high-dimensional model is: , in, This is the aerodynamic performance response function. For zero-order component terms, The total number of parameter groups, Indicates the first Class parameter group, For the first Items within a class parameter group, For the first Class and the Inter-group items between class parameter groups.

[0012] Specifically, within-group items Further expressed as: , Between-group terms Further expressed as: , in, For the first The number of parameters in the class parameter group For the first The first in the class parameter group The independent action components of each parameter, For the first The first in the class parameter group The and the first Intra-group interaction components between parameters For the first The first in the class The parameter and the first The first in the class Intergroup interaction components among parameters.

[0013] Specifically, the aerodynamic uncertainty analysis of the compressor includes: Calculate the contribution rate of each parameter category to the mean aerodynamic performance index; Calculate the contribution rate of each parameter category to the variance of aerodynamic performance index fluctuation; Calculate the intergroup interaction index of the influence of coupling effects between different parameter categories on aerodynamic performance indicators.

[0014] Specifically, the analysis results include the ranking of parameter categories by importance and the results of interactive identification of key parameter categories.

[0015] Specifically, it also includes the following steps: Based on the importance ranking of parameter categories in the analysis results, identify the dominant parameter categories that cause fluctuations in compressor aerodynamic performance; Based on the key parameter category interaction identification results in the analysis results, the coupling mechanism between different parameter categories is determined; Based on the dominant parameter categories and coupling mechanisms, a compressor manufacturing tolerance allocation scheme, a geometric optimization design scheme, or an operating condition evaluation report is generated.

[0016] The technical solution provided by this invention has the following advantages compared with the prior art: This invention represents the NHDMR format through a nested high-dimensional model, innovatively reorganizing the traditional HDMR components into nested, mathematically equivalent but physically meaningful groups. This allows for the precise quantification of the individual and coupled effects of multiple parameters and multiple categories, transforming the surrogate model from a black box to a white box, greatly enhancing interpretability. It not only enables the categorical and hierarchical analysis of compressor uncertainty factors but also facilitates multi-level sensitivity and correlation analysis in aerodynamic uncertainty quantification problems involving multiple categories of factors. It can analyze compressor aerodynamic uncertainties at the parameter category level, improving the engineering interpretability of the results; it can quantitatively identify the impact of different parameter categories and their interactions on aerodynamic performance; it helps to clarify the main sources of compressor performance fluctuations, providing a basis for decision-making in engineering design and manufacturing; and it is applicable to compressor aerodynamic analysis problems with high-dimensional, multi-category uncertainty parameters. Attached Figure Description

[0017] Figure 1 A flowchart of a compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation provided by an embodiment of the present invention; Figure 2 This invention provides a category influence distribution diagram after classifying uncertainty parameters according to geometric dimensions; Figure 3 This is a category influence distribution diagram after classifying uncertainty parameters according to blade type, provided as an embodiment of the present invention. Detailed Implementation

[0018] The following detailed description of a specific embodiment of the present invention is provided in conjunction with the accompanying drawings. However, it should be understood that the scope of protection of the present invention is not limited to the specific embodiment.

[0019] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the technical solution of this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0020] The present invention will be described below through several specific embodiments. To keep the following description of the embodiments clear and concise, detailed descriptions of known functions and components may be omitted. When any component of an embodiment of the present invention appears in more than one drawing, the component may be represented by the same reference numerals in each drawing.

[0021] Figure 1 This invention provides a flowchart of a compressor aerodynamic uncertainty analysis method based on a nested high-dimensional model representation. Figure 2 This invention provides a category influence distribution diagram after classifying uncertainty parameters according to geometric dimensions. Figure 3 This is a category influence distribution diagram after classifying uncertainty parameters according to blade type, provided as an embodiment of the present invention.

[0022] like Figure 1 As shown, this invention provides a method for compressor aerodynamic uncertainty analysis based on a nested high-dimensional model representation, comprising the following steps: acquiring and classifying the compressor's uncertainty parameters, dividing the uncertainty parameters into multiple parameter categories with different engineering physical meanings; constructing the compressor aerodynamic response function to establish a mapping relationship between the uncertainty parameters and the compressor aerodynamic performance indicators, and decomposing the mapping relationship based on the high-dimensional model representation method to obtain multiple component functions; according to the parameter category classification, performing nested recombination of the decomposed component functions to construct a nested high-dimensional model representation model; inputting the uncertainty parameters into the nested high-dimensional model representation model to perform compressor aerodynamic uncertainty analysis; and outputting the analysis results, which are used to guide the compressor's manufacturing tolerance allocation, geometric optimization design, or operating condition evaluation.

[0023] Specifically, the mapping relationship between uncertain parameters and compressor aerodynamic performance indicators is established, and the mapping relationship is decomposed based on a high-dimensional model representation method, including: , in, For zero-order component terms, It is a first-order component term. Representing variables and The second-order component term generated by the coupling of related effects after removing the influence of individual effects.

[0024] Specifically, the uncertain parameters of the compressor are obtained, and the uncertain parameters are used as input variables. The aerodynamic performance indicators of the compressor are used as output responses. The aerodynamic performance indicators include, but are not limited to, indicators such as adiabatic efficiency and total pressure ratio. Numerical simulations are used to obtain aerodynamic performance index samples corresponding to different parameter changes, forming an input-output sample set. Based on HDMR, the output response is decomposed to construct zero-order, first-order, and higher-order interaction terms. Each component function is obtained through regression fitting or interpolation fitting, thus forming a surrogate mapping model from parameters to aerodynamic performance indices. According to the concept of HDMR, input variables... With the corresponding output There exists a proxy model form as shown in equation (1). It is a zero-order component term, which is a constant. It is a first-order component term, which quantifies the variable. right The impact. Representing variables and The second-order component term is generated by the coupling of related effects after removing the influence of individual effects. The subsequent term is the higher-order component term, which represents the coupling effect when more input variables act together on the output after excluding the influence of low-order correlations. Since the second-order HDMR (Equation (2)) model can characterize the individual effects and pairwise coupling effects of input variables and achieve a good balance between modeling accuracy and computational complexity, the second-order HDMR model is used as an example for subsequent analysis.

[0025] (1), Furthermore, to clarify the specific objects being processed and their classification criteria, this embodiment of the invention provides specific limitations on the uncertainty parameters. It specifies that the parameters must include at least one of the following: geometric dimensions, profile control, angle, material properties, incoming flow conditions, and operating condition parameters. It also provides several engineering-feasible classification methods, such as by source, dimension, region, or blade type. This limitation makes the application objects of this method clearer, the parameter classification more operable, and lays the logical foundation for subsequent nested recombination.

[0026] Specifically, uncertain parameters include geometric dimension parameters, profile control parameters, angle parameters, material property parameters, incoming flow condition parameters, and operating condition parameters; the parameter categories can be divided by: uncertainty source, geometric dimension, structural region, aerodynamic component, or blade type.

[0027] Furthermore, geometric uncertainty parameters include, for example, compressor inlet total temperature, total pressure, and inlet distortion; compressor operating speed; impeller or blade geometric dimensions, profile control parameters, and angle parameters; impeller or blade material properties; and machining deviations in the stator region. For any specific research task, based on the engineering physics meaning, research objectives, or analysis task, a research perspective is first determined. From this perspective, the geometric uncertainty parameters are divided into multiple categories to characterize the relationships between different categories under the perspective of interest to this task. Each category corresponds to a parameter group, and each parameter group contains multiple parameters. The parameter sets include, but are not limited to: when studying the influence of uncertain parameters from different sources, they can be classified into categories such as incoming flow, geometry, operating conditions, or materials; when studying the response differences of structural components or variable regions in compressor uncertainties, they can be classified into categories such as moving blades, stationary blades, or different stages; when studying different dimensions or dimensionality reduction problems, they can be classified into one-dimensional, two-dimensional, and three-dimensional parameter sets according to geometric dimensions; when studying the influence of local structural regions, they can be classified into hub-side and tip-side parameter sets according to structural regions; when studying the differences in the effects of main blades and splitter blades, they can be classified into main blade parameter sets and splitter blade parameter sets according to blade type, etc.

[0028] If a parameter is incorrectly categorized into a group that does not match its physical properties, structural location, or the component it belongs to, it is considered improper classification, which may lead to inaccurate physical meaning of the parameter group and consequently distorted analytical results. Since the classification is performed after the research perspective is determined, the classification criteria are usually clear and verifiable. If a parameter is categorized into a group that does not match its physical meaning, the component it belongs to, its structural location, or its geometric dimensions, it can be considered improperly classified. For example, when classifying by moving blade / stationary blade, moving blade parameters will not be classified into the stationary blade parameter group; when classifying by hub side / blade tip side, blade tip parameters will not be classified into the hub side parameter group. Therefore, when the classification is correct, the parameter grouping can accurately represent the parameter relationships under the corresponding perspective; only when the classification criteria under that perspective are violated will the analytical results be distorted.

[0029] Furthermore, to clarify the characterization method of aerodynamic performance and the construction approach of the response function, this embodiment of the invention refines the aerodynamic performance indicators and aerodynamic response functions. It specifies the performance indicators as the core adiabatic efficiency or total pressure ratio in the compressor field, and clarifies the construction method of the response function as a combination of numerical simulation and high-dimensional model representation methods. This limitation ensures that the method can quantitatively analyze the key performance of compressors and provides specific and feasible technical implementation means.

[0030] Specifically, aerodynamic performance indicators include: adiabatic efficiency, isentropic efficiency, total pressure ratio, static pressure ratio, flow rate, power, power consumption, surge margin, blockage margin, stable operating range, loss coefficient, pressure recovery coefficient, and other indicators characterizing compressor performance; the aerodynamic response function is constructed by combining numerical simulation or experimental testing with a high-dimensional model.

[0031] Furthermore, to clearly define the mathematical implementation of nested recombination, this invention provides a depth-limited representation of the nested high-dimensional model. The second-order expansion of the model is given, revealing macroscopically that the aerodynamic response can be decomposed into independent contributions from each parameter category and interactive contributions between different categories. Further, the internal mathematical structures of "within-group terms" and "between-group terms" are dissected, clarifying that they are composed of the independent effects of parameters within a category, the interactive effects between parameters within a category, and the interactive effects between parameters across categories. Together, these mathematically solve the core technical problem of how to reconstruct the traditional "single-parameter" analysis framework into a "multi-level, structured" analysis framework with engineering significance, which is the mathematical foundation for the engineering interpretability of this method.

[0032] Specifically, nested recombination is as follows: Let there be a total of The parameter category, the first The categories include Given several uncertain parameters, the second-order expansion of the nested high-dimensional model is: , in, This is the aerodynamic performance response function. For zero-order component terms, The total number of parameter groups, Indicates the first Class parameter group, For the first Items within a class parameter group, For the first Class and the Inter-group items between class parameter groups.

[0033] Specifically, the compressor aerodynamic response function is constructed as follows: (2), Based on the HDMR method, the mapping relationship is decomposed; according to the parameter grouping method, the component functions in the high-dimensional model representation are nested and reorganized to construct the NHDMR model. To distinguish between the standard HDMR and the proposed NHDMR scheme, the second-order HDMR extension is first expressed as equation (3), where Represents the vector of uncertain factors, by It consists of several components.

[0034] (3), NHDMR organizes its data into a multi-level structure based on the categories of uncertainty factors. Let... This represents the total number of uncertain categories. Indicates the first The number of uncertain factors, i.e. At the same time, set , .assumed , Then the second-order expansion of the NHDMR model is given by equations (4)-(6). For a single-class component term in NHDMR expansion, it represents the first... Uncertainty factors The effect on the output response when used alone. There are two types of interactive items, reflecting and removing categories. and The coupling effect of each individual action on the output response. refer to Category No. Each component The effect of acting independently on the output response. express Category No. and the The interactions between the components after removing their individual effects. Similarly, Indicated Category No. Each component and Category No. The interaction between the components after removing their individual effects.

[0035] (4), (5), (6), Therefore, the model expression can clearly distinguish the independent contributions of each parameter category to aerodynamic performance. This includes individual contributions within the parameter category. and interactive contributions The combined influence; the interactive contributions between different parameter categories .

[0036] Furthermore, to clarify the specific content of uncertainty analysis, this embodiment of the invention specifies the aerodynamic uncertainty analysis of a compressor. It defines the analysis process as including at least calculating the contribution rate of each parameter category to the performance mean, the contribution rate to the performance fluctuation variance, and the inter-group interaction index between different categories. This definition transforms the general concept of "analysis" into specific, quantifiable, and calculable indicators.

[0037] Specifically, the aerodynamic uncertainty analysis of the compressor includes: calculating the contribution rate of each parameter category to the mean of aerodynamic performance indicators; calculating the contribution rate of each parameter category to the variance of aerodynamic performance indicators; and calculating the intergroup interaction index of the coupling effect between different parameter categories on aerodynamic performance indicators.

[0038] Furthermore, based on the nested high-dimensional model representation, uncertainty analysis of the compressor's aerodynamic performance can be performed. This analysis not only reveals the influence of individual parameters on the statistical characteristics of aerodynamic performance and the degree of coupling between different parameters in traditional HDMR, but also the influence of each category of parameters on the aerodynamic performance mean, fluctuation (standard deviation), and the coupling between different categories of parameters. For example, Figure 2 This represents the distribution of category effects after classifying the uncertainty parameters according to geometric dimensions. Here, 1D represents a one-dimensional parameter group, 2D / 3D represents two-dimensional and three-dimensional parameter groups, Coupling Effect represents the coupling effect between different parameter groups, and SUM represents the combined result of the effects of each category. Figure 2 This is used to characterize the individual and coupled effects of different parameter groups on the output response compressor adiabatic efficiency φ from a geometric perspective. Figure 3 This is a distribution diagram of the category effects of uncertain parameters after classification by blade type. Here, Main Blade represents the main blade parameter group, Splitter Blade represents the splitter blade parameter group, Others represents other parameter groups besides the main and splitter blades, Coupling Effect represents the coupling effect between different parameter groups, and SUM represents the combined result of the effects of each category. Figure 3 This is used to characterize the individual and coupled effects of different parameter groups on the output response compressor adiabatic efficiency φ from the perspective of blade type.

[0039] Furthermore, to clarify the specific presentation format of the analysis results, this embodiment of the invention refines the output analysis results. It concretizes the analysis results into a ranking of parameter category importance and an interactive identification result of key parameter categories. This limitation transforms complex mathematical and statistical results into decision-making basis that engineers can directly understand and use.

[0040] Specifically, the analysis results include the ranking of parameter categories by importance and the results of interactive identification of key parameter categories.

[0041] The output provides analytical results for identifying sources of uncertainty in the compressor. These results include a ranking of parameter categories based on their importance and the results of cross-identification of key parameters. The analytical results can be used to guide the allocation of compressor manufacturing tolerances, geometric optimization design, or operational condition assessment.

[0042] Furthermore, to fully demonstrate the engineering application value of this invention, the embodiments of this invention add an engineering decision-making application step based on the above steps. This defines how to utilize the aforementioned analysis results to generate specific manufacturing tolerance allocation schemes, geometric optimization design schemes, or operational condition evaluation reports. This limitation elevates the method from a simple analysis tool to a complete solution that can directly guide engineering design, manufacturing, and operation.

[0043] Specifically, the process also includes the following steps: identifying the dominant parameter categories that cause fluctuations in compressor aerodynamic performance based on the importance ranking of parameter categories in the analysis results; determining the coupling mechanism between different parameter categories based on the key parameter category interaction identification results in the analysis results; and generating compressor manufacturing tolerance allocation schemes, geometric optimization design schemes, or operating condition evaluation reports based on the dominant parameter categories and coupling mechanisms.

[0044] The above inventions are merely a few specific embodiments of the present invention. However, the embodiments of the present invention are not limited thereto, and any variations that can be conceived by those skilled in the art should fall within the protection scope of the present invention.

Claims

1. A method for analyzing compressor aerodynamic uncertainties based on nested high-dimensional model representation, characterized in that, Includes the following steps: Acquire and classify the uncertainty parameters of the compressor, and divide the uncertainty parameters into multiple parameter categories with different engineering physical meanings; A compressor aerodynamic response function is constructed to establish a mapping relationship between the uncertain parameters and the compressor aerodynamic performance indicators. The mapping relationship is then decomposed based on a high-dimensional model representation method to obtain multiple component functions. Based on the parameter categories, the component functions obtained from the decomposition are reorganized in a nested manner to construct a nested high-dimensional model representation model; The uncertainty parameters are input into the nested high-dimensional model representation model to perform compressor aerodynamic uncertainty analysis; The analysis results are output and used to guide the allocation of manufacturing tolerances, geometric optimization design, or operational condition evaluation of the compressor.

2. The compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation as described in claim 1, characterized in that, The establishment of the mapping relationship between the uncertainty parameters and the compressor aerodynamic performance indicators, and the decomposition of the mapping relationship based on a high-dimensional model representation method, specifically includes: ， in, For zero-order component terms, It is a first-order component term. Representing variables and The second-order component term generated by the coupling of related effects after removing the influence of individual effects.

3. The compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation as described in claim 1, characterized in that, The uncertain parameters include geometric dimension parameters, profile control parameters, angle parameters, material property parameters, incoming flow condition parameters, and operating condition parameters; The parameter categories can be classified in the following ways: by source of uncertainty, by geometric dimension, by structural region, by aerodynamic component, or by blade type.

4. The compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation as described in claim 1, characterized in that, The aerodynamic performance indicators include: adiabatic efficiency, isentropic efficiency, total pressure ratio, static pressure ratio, flow rate, power, power consumption, surge margin, blockage margin, stable operating range, loss coefficient, pressure recovery coefficient, and other indicators characterizing compressor performance; the aerodynamic response function is constructed through numerical simulation or experimental testing combined with high-dimensional model representation methods.

5. The compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation as described in claim 1, characterized in that, The nested recombination specifically refers to: Let there be a total of The parameter category, the first The categories include Given several uncertain parameters, the second-order expansion of the nested high-dimensional model is: ， in, This is the aerodynamic performance response function. For zero-order component terms, The total number of parameter groups, Indicates the first Class parameter group, For the first Items within a class parameter group, For the first Class and the Inter-group items between class parameter groups.

6. The compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation as described in claim 5, characterized in that, The items within the group Further expressed as: ， The inter-group items Further expressed as: ， in, For the first The number of parameters in the class parameter group For the first The first in the class parameter group The independent action components of each parameter, For the first The first in the class parameter group The and the first Intra-group interaction components between parameters For the first The first in the class The parameter and the first The first in the class Intergroup interaction components among parameters.

7. The compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation as described in claim 1, characterized in that, The compressor aerodynamic uncertainty analysis includes: Calculate the contribution rate of each parameter category to the mean aerodynamic performance index; Calculate the contribution rate of each parameter category to the variance of aerodynamic performance index fluctuation; Calculate the intergroup interaction index of the influence of coupling effects between different parameter categories on aerodynamic performance indicators.

8. The compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation as described in claim 1, characterized in that, The analysis results include the ranking of parameter categories by importance and the results of interactive identification of key parameter categories.

9. The compressor aerodynamic uncertainty analysis method based on nested high-dimensional model representation as described in claim 1, characterized in that, It also includes the following steps: Based on the importance ranking of the parameter categories in the analysis results, identify the dominant parameter categories that cause fluctuations in the compressor's aerodynamic performance; Based on the key parameter category interaction identification results in the analysis results, the coupling mechanism between different parameter categories is determined; Based on the dominant parameter categories and coupling mechanisms, a compressor manufacturing tolerance allocation scheme, a geometric optimization design scheme, or an operating condition evaluation report is generated.

Images